Estimation of Vehicle Roll Angle

Sammanfattning: Haldex Traction can with their Active Yaw Control prevent unwanted handling by applying an extra yaw torque with their all wheel drive system. To be able to calculate when or when not, yaw torque should be applied, it is important to accurate know the different state information about the vehicle. When driving on banked roads, vehicle dynamics and sensor measurements are changed compared to driving on a flat surface. Because of this it is desirable to know the degree of banking the vehicle is exposed to. The estimation of the banking is made with sensors already present in modern production cars; lateral accelerometer, yaw rate gyro, steering wheel angle and longitudinal velocity. This by isolating the part of the measured lateral acceleration that is derived from the normal forces due to gravity. To be able to make a good and stable estimation it is necessary to also estimate the vehicle's lateral velocity, and especially its derivative. This is done by estimating vehicle states with a single track bicycle model. The model has been used in former thesis works, but it was in this thesis extended with the angle of the road as a parameter. Two different observers have been evaluated for measurement update of the model; Extended Kalman filter (EKF) and an Averaging observer. Evaluation of the algorithms have been done in Haldex simulator VehSim running in Matlab/Simulink and real measurements from a test vehicle.